Churn Rate Formula

customer churn % = customers lost / customers at start × 100
revenue churn %  = MRR lost / MRR at start × 100
average lifetime = 100 / customer churn %

Revenue churn counts downgrades as lost dollars as well as cancellations. That is deliberate: a customer who drops from a $500 plan to a $200 plan removes $300 of recurring revenue exactly like a cancellation of $300 would.

Churn is a rate applied to the population that was actually at risk. Those are the customers present at the start. Three denominators are in common use:

• Start of period — the standard, and what published benchmark tables mean. Use this.
• End of period — the base is smaller after the losses, so the rate is overstated. It flatters nothing, but it is not comparable.
• Average of start and end — understates churn in a fast-growing business, because the average is pulled up by customers acquired during the period who were never at risk.

If you report churn to investors or compare it with a benchmark, state the denominator. A one-point difference is easily explained by this choice alone.

You start the month with 500 customers and 15 cancel:

churn = 15 / 500 × 100 = 3%
lifetime = 100 / 3 ≈ 33 months

If those 500 customers represent $50,000 of MRR and you lose $2,000 to cancellations plus $500 to downgrades, revenue churn is $2,500 / $50,000 = 5%. The business is losing fewer logos than dollars — the reverse of the healthy pattern — which usually means larger accounts are leaving.

Multiplying a monthly rate by twelve overstates annual churn badly. At 5% monthly:

annual retention = (1 − 0.05)^12 = 0.54
annual churn     = 46%

Not 60%. Churn removes a slice of whatever remains, not a slice of the original base. Going the other way, an annual churn of 46% is about 5% monthly, not 3.8%.

Put opening customers in B2, customers lost in C2, opening MRR in D2 and MRR lost in E2:

customer churn % = C2 / B2 * 100
revenue churn %  = E2 / D2 * 100
average lifetime = 100 / (C2 / B2 * 100)
12-month retention = (1 - C2 / B2) ^ 12

The last line is the one people get wrong: exponentiate, do not multiply by twelve.

Report both, every month. Their relationship is the signal:

• Revenue churn below logo churn — you keep the customers who pay most. Healthy, and if expansion exceeds the losses you have negative net revenue churn.
• Revenue churn above logo churn — you are losing fewer accounts but more money, which almost always means the largest accounts are leaving. Materially worse than the logo count suggests.

Expansion revenue should be reported alongside churn, not netted into it silently. Negative net revenue churn means the existing base would grow even with zero new customers.

These bands come from ranges published by OpenView, ChartMogul and ProfitWell. Segment by plan and channel before comparing: a blended rate averages a 1% enterprise cohort with a 9% self-serve cohort and describes neither.

Retention is usually cheaper to fix than acquisition is to buy. The highest-leverage levers in most early-stage businesses are:

• Involuntary churn. Failed payments are often 20%–40% of total churn in self-serve products and are almost entirely fixable with dunning emails and card-updater services.
• Onboarding. Churn is highest in the first 90 days; if cohort curves never flatten, the problem is activation, not the product.
• Annual plans. They convert a monthly decision to leave into a yearly one, and they improve cash collection at the same time.
• Segmentation. A single blended rate hides which plan, channel or cohort is leaking.

Run the numbers on the churn calculator, which includes a twelve-month compounding projection, or see how churn flows into lifetime value in the LTV:CAC guide.